Aberrantly methylated DNA is found abundantly in all forms of cancer and the tumors that produce them. In fact, it has been estimated that within the cells of every tumor are several hundreds of aberrantly methylated CpG islands, many of which are promoters of tumor suppressor genes. The ability to detect and quantify promoter methylation will allow much more refined diagnostic and prognostic stratification. Recently, several groups including ours have reported the detection of tumor-associated methylated DNA circulating in serum/plasma. The use of circulating methylated DNA is a particularly attractive for screening and companion diagnostics for cancer, as serum/plasma is obtained through a simple, relatively noninvasive procedure. Currently, detection of circulating methylated DNA is mainly performed using bisulfite-based methods such as methylation specific PCR (MSP) due to their high sensitivity and specificity. However, clinical applications of these tests have been encumbered by a number of hurdles, resulting in highly variable success. For any bisulfite based methods, the process of DNA extraction and bisulfite conversion involves several disconnected steps on different platforms, resulting in substantial sample loss. Furthermore, MSP is designed to detect a specific methylation pattern; however, the promoter methylation patterns may be highly variable in tumors, compromising its clinical sensitivity. While sequence-based methods including bisulfite sequencing and pyrosequencing can be used to analyze methylation heterogeneity, these methods do not have the requisite sensitivity to detect the extremely low ratios (<0.1%) of methylated epialleles present in the bloodstream. In order to address these issues, we propose a streamlined methylation detection platform combining a silica paramagnetic bead (SSB)-based method for processing circulating DNA from large volume plasma samples to maximize assay fidelity and a microfluidic digital high resolution melt (HRM) approach for detecting and quantifying heterogeneous promoter epialleles at the single molecule level. A microfluidic device will be developed to facilitate high fidelity digital PCR and HRM in 1.6X10^6 microfluidic reaction chambers. A melt curve analysis program will be developed for analyzing the specific methylation allele in each reaction chamber according to the respective melt profile. The platform will be validated using both synthetic control samples and clinical samples from lung cancer patients undergoing epigenetic therapy. The proposed technology will enable detection and quantification of heterogeneous methylation with a low LOD (<10 copies of methylated DNA in e 2 ml plasma sample), high sensitivity (1/100,000; methylated/unmethylated alleles) and a wide dynamic range (7 orders of magnitude), a level of performance unattainable by any other existing technologies (MSP, real- time qMSP and sequencing).